Glossary: Score
Score is Jev’s ordered-rubric primitive on TypeSafe System One. You send type: "score" with criteria as an array of 2–10 level descriptions. The answer includes score (a probability-weighted position that may sit between levels), legend, probabilities, and confidence. The score may be fractional — it is an expectation, not an enum index. Jev does not generate a rationale paragraph.
This glossary entry is for builders who already heard of Jev / System One and need a citable definition before they interpolate a Score into dollars. jev.pro is an independent unofficial field guide — not affiliated with TypeSafe AI. Confirm the primitive on docs.typesafe.ai. We do not sell or issue API keys.
How it shows up
Write levels as descriptions a person could apply (“Calm”, “Frustrated”, “Very angry”), not naked integers. All questions in one POST /v1/systemone share the same state. Question IDs are not used in inference. Do not treat a fractional score as a physical quantity — TypeSafe’s jaggedness note says jev-1.13 is weak at numerical calibration. Threshold “past level 1”; do not reconstruct a dollar amount.
Split multi-factor scores and keep the arithmetic in code. Ticket priority that mixes severity, frustration, and engineer-readiness is three Scores, then weights you own. Asking Jev to multiply weights is the wrong contract. Pin jev-1.13.0 (or the current versioned id) once you fit a threshold on the expectation or on confidence. jev-latest can move.
confidence is derived from the Score probabilities map, same family as Choice. It is a group-level calibration story under the RLCD vendor claim — not a per-row guarantee. Schema-safe ≠ factually correct: a valid rubric position can still be the wrong read of the ticket.
Contrast
Score vs Choice. Choice is unordered classification (max 255 options). Score is an ordered rubric (2–10 levels). Use Choice for “which team?” Use Score for “how much?”
Score vs Noul. Noul is P(yes) in [0, 1] with no confidence and no probabilities map. A Noul of 0.5 is not a “medium Score”; yes and no are equally likely.
Score vs a threshold. The score is position on your rubric. A threshold is the cut your code applies to that position, to confidence, or to a Noul. TypeSafe examples are illustrations, not defaults.
Score vs composite “priority.” One Score should be one judgment. Composition lives in your service.
Common mistakes
- Asking Jev to multiply or average weights. Keep math in code.
- Using the expectation to reconstruct a magnitude between two levels.
- Treating 1.05 on a 0–2 rubric as illegal. It is a legal expectation.
- Reading Noul 0.5 as a Score midpoint.
- Forgetting Score has
confidence; Noul does not.
See also
FAQ
Can I ask Jev to multiply weights?
No. Keep arithmetic in code. Split multi-factor scores into atomic Scores, then weight them yourself.
Does Score have confidence?
Yes. Noul does not.
Can score be 1.05 on a 0–2 rubric?
Yes — it is an expectation, not an enum index.
Should I threshold score or confidence?
Often both when stakes are high. score is “how far along the rubric.” confidence is “how peaked the distribution is.”
Limits
This is a jev.pro working definition. Score is a TypeSafe API question type. We do not invent search volume. Schema-safe answers are not automatically factually correct.
Disclaimer
This is an independent unofficial site and is not affiliated with TypeSafe AI; official documentation is available at https://docs.typesafe.ai. Never treat jev.pro as TypeSafe official documentation. We do not sell, issue, or proxy API keys.